Metal Artifact Correction and Noise-Adjustment of CT Scan Image

1. A method for correcting a CT scan image, the method comprising: extracting a target image including a target region to be corrected from an original CT scan image; obtaining a target projection range corresponding to the target region by orthographically projecting the target image, and correcting target projection data within the target projection range to obtain corrected target projection data; generating noise-adjusted target projection data by adjusting noise of the corrected target projection data according to target projection data of an intensity domain which is obtained from target projection data of an attenuation domain, so as to improve noise consistency between the target region and rest of the CT original scan image; and reconstructing a corrected CT scan image based on the noise-adjusted target projection data; wherein adjusting noise of the corrected target projection data to generate the noise-adjusted target projection data comprises: converting the target projection data of the attenuation domain into the target projection data of the intensity domain; multiplying the target projection data of the intensity domain by a default transform coefficient so as to obtain a transformed target projection data; using the transformed target projection data as a Poisson parameter to calculate a Poisson random number, and determining update transformed target projection data based on the Poisson random number; transforming the update transformed target projection data to update target projection data of the intensity domain; converting the update target projection data of the intensity domain into update target projection data of the attenuation domain; and determining the update target projection data of the attenuation domain as the noise-adjusted target projection data.

2. The method of claim 1 , wherein obtaining a target projection range corresponding to the target region by orthographically projecting the target image, and correcting target projection data within the target projection range comprises: obtaining a linear interpolation value of two projection data points located at two adjacent sides of the target projection data; replacing the target projection data with the linear interpolation value so as to obtain corrected data of the target projection data; reconstructing a corresponding corrected image according to the corrected data of the target projection data; determining a reference image of the corrected image according to a default segmentation threshold; orthographically projecting the reference image so as to obtain reference projection data of the reference image; and determining the reference projection data as the corrected target projection data.

3. The method of claim 1 , wherein reconstructing a corrected CT scan image based on the noise-adjusted target projection data comprises: generating a first corrected CT scan image based on the noise-adjusted target projection data; and integrating the first corrected CT scan image with the target image so as to generate a second corrected CT scan image.

4. The method of claim 1 , wherein the target region to be corrected is a metal region.

5. The method of claim 1 , wherein determining update transformed target projection data based on the Poisson random number includes using the Poisson random number as the update transformed target projection data.

6. A device for correcting a CT scan image, the device comprising: a processor configured to execute machine readable instructions corresponding to a control logic for correcting a CT scan image stored on a storage medium, wherein the machine readable instructions, when executed by the processor, cause the processor to: extract a target image including a target region to be corrected from an original CT scan image; obtain a target projection range corresponding to the target region by orthographically projecting the target image, and correct target projection data within the target projection range to obtain corrected target projection data; generate noise-adjusted target projection data by adjusting noise of the corrected target projection data according to target projection data of an intensity domain which is obtained from target projection data of an attenuation domain, so as to generate noise-adjusted target projection data to improve noise consistency between the target region and rest of the CT original scan image; and reconstruct a corrected CT scan image based on the noise-adjusted target projection data wherein when adjusting noise of the corrected target projection data to generate the noise-adjusted target projection data, said machine readable instructions further cause the processor to: convert the target projection data of the attenuation domain into the target projection data of the intensity domain; multiply the target projection data of the intensity domain by a default transform coefficient so as to obtain transformed target projection data; use the transformed target projection data as a Poisson parameter to calculate a Poisson random number, and determine update transformed target projection data based on the Poisson random number; transform the update converted target projection data to update target projection data of the intensity domain; convert the update target projection data of the intensity domain into update target projection data of the attenuation domain; and determine the update target projection data of the attenuation domain as the noise-adjusted target projection data.

7. The device according to claim 6 , wherein said machine readable instructions further cause the processor to: obtain a linear interpolation value of two projection data points located at two adjacent sides of the target projection data; replace the target projection data with the linear interpolation value so as to obtain corrected data of the target projection data; reconstruct a corresponding corrected image according to the corrected data of the target projection data; determine a reference image of the corrected image according to a default segmentation threshold; orthographically project the reference image so as to obtain reference projection data of the reference image; and determine the reference projection data as the corrected target projection data.

8. The device according to claim 6 , wherein said machine readable instructions further cause the processor to: generate a first corrected CT scan image based on the noise-adjusted target projection data; and integrate the first corrected CT scan image with the target image so as to generate a second corrected CT scan image.

9. The device according to claim 6 , wherein the target region to be corrected is a metal region.

10. The device of claim 6 , wherein determining update transformed target projection data based on the Poisson random number includes using the Poisson random number as the update transformed target projection data.